Defining brain circuits that control decisions, such as those between social interactions and asocial behaviors is an important problem in neuroscience with high relevance to human health. These neural circuits are located in evolutionarily ancient brain regions, such as the amygdala. In humans, decreased social interactions are a key symptom domain in psychiatric disorders such as autism, and are thought to involve the amygdala. The amygdala is a complex structure consisting of at least 12 distinct subregions. The amygdala circuits that control conditioned fear in the basolateral and central amygdala have been intensively studied. However a gap remains between our understanding of these circuits, and those in different amygdala subnuclei that control social interactions. The latter are thought to be located in the medial subdivision of the amygdala (MeA). To fill this gap, we will begin to dissect the function of MeA circuits that control the balance between social interactions and repetitive self-grooming, an asocial behavior. This balance is important because disorders such as autism often feature increased repetitive asocial behaviors, as well as decreased social interactions. The long-term goal is to understand the circuit-level control of this balance at a brain-wide level. The overall objective of this application is to define the roleof different MeA neuronal subpopulations that antagonistically control social interactions vs. repetitive self-grooming, and to understand the circuitry through which this antagonism is exerted. The central objective of this proposal is to study how GABAergic and glutamatergic neuronal subpopulations in the medial amygdala reciprocally regulate these opponent activities. The rationale for this research is that it will reveal fundamental mechanisms of neural circuit function in a brain region that is relevant to human health. To achieve our objective, we will map the functional projections of vGAT+ and VGLUT2+ MeApd neurons that control social interactions vs. repetitive self-grooming, respectively (Aim 1); test the hypothesis that social interactions are controlled by a dis-inhibition circuit and map that circuit (Aim 2); determine the mechanism by which vGAT+ and vGLUT2+ subpopulations exert antagonistic control of social vs. self-grooming behaviors (Aim 3); determine whether the dual functions performed by each of these subpopulations derive from common or distinct cell types (Aim 4). The contribution will be to apply state-of-the- art genetically based tools to dissect circuit-level mechanisms in MeA that control social vs. repetitive asocial behaviors. This contribution is significant because it will oen up the study of amygdala circuitry controlling social interactions, at a level of cellular specificty that has not yet been achieved. The contribution is innovative, because it investigates novel features of amygdala circuitry that we have recently uncovered involving the excitation:inhibition balance. The work proposed in this application will therefore both advance our basic understanding of neural circuit functional organization, and shed light on the particulars of amygdala neuronal subpopulations and circuitry that may relevant to human disorders affecting social interactions.